Analyzing color imaging failure on consumer-grade cameras.

There are many efforts to employ consumer-grade cameras for home-based health and wellness monitoring. Such applications rely on users to capture images for analysis using their personal cameras in a home environment. When color is a primary feature for diagnostic algorithms, the camera requires calibration to ensure accurate color measurements. Given the importance of these diagnostic tests for the users' health and well-being, it is important to understand the conditions in which color calibration may fail. To this end, we analyzed a wide range of camera sensors and environmental lighting to determine (1) how often color calibration failure is likely to occur and (2) the underlying reasons for failure. Our analysis shows that it is rare to encounter a camera sensor and lighting condition combination that results in color imaging failure. Moreover, when color imaging does fail, the cause is almost always attributed to spectral poor environmental lighting and not the camera sensor. We believe this finding is useful for scientists and engineers developing color-based applications for use with consumer-grade cameras.

Unexpected role for IGF-1 in starvation: Maintenance of blood glucose.

Wild-type (WT) mice maintain viable levels of blood glucose even when adipose stores are depleted by 6 d of 60% calorie restriction followed by a 23-h fast (hereafter designated as "starved" mice). Survival depends on ghrelin, an octanoylated peptide hormone. Mice that lack ghrelin suffer lethal hypoglycemia when subjected to the same starvation regimen. Ghrelin is known to stimulate secretion of growth hormone (GH), which in turn stimulates secretion of IGF-1 (insulin-like growth factor-1). In the current study, we found that starved ghrelin-deficient mice had a 90% reduction in plasma IGF-1 when compared with starved WT mice. Injection of IGF-1 in starved ghrelin-deficient mice caused a twofold increase in glucose production and raised blood glucose to levels seen in starved WT mice. Increased glucose production was accompanied by increases in plasma glycerol, fatty acids and ketone bodies, and hepatic triglycerides. All of these increases were abolished when the mice were treated with atglistatin, an inhibitor of adipose tissue triglyceride lipase. We conclude that IGF-1 stimulates adipose tissue lipolysis in starved mice and that this lipolysis supplies energy and substrates that restore hepatic gluconeogenesis. This action of IGF-1 in starved mice is in contrast to its known action in inhibiting adipose tissue lipase in fed mice. Surprisingly, the ghrelin-dependent maintenance of plasma IGF-1 in starved mice was not mediated by GH. Direct injection of GH into starved ghrelin-deficient mice failed to increase plasma IGF-1. These data call attention to an unsuspected role of IGF-1 in the adaptation to starvation.

Interplay between Asters/GRAMD1s and phosphatidylserine in intermembrane transport of LDL cholesterol.

Low-density lipoprotein (LDL) delivers cholesterol to mammalian cells through receptor-mediated endocytosis. The LDL cholesterol is liberated in lysosomes and transported to the plasma membrane (PM) and from there to the endoplasmic reticulum (ER). Excess ER cholesterol is esterified with a fatty acid for storage as cholesteryl esters. Recently, we showed that PM-to-ER transport of LDL cholesterol requires phosphatidylserine (PS). Others showed that PM-to-ER transport of cholesterol derived from other sources requires Asters (also called GRAMD1s), a family of three ER proteins that bridge between the ER and PM by binding to PS. Here, we use a cholesterol esterification assay and other measures of ER cholesterol delivery to demonstrate that Asters participate in PM-to-ER transport of LDL cholesterol in Chinese hamster ovary cells. Knockout of the gene encoding PTDSS1, the major PS-synthesizing enzyme, lowered LDL-stimulated cholesterol esterification by 85%, whereas knockout of all three Aster genes lowered esterification by 65%. The reduction was even greater (94%) when the genes encoding PTDSS1 and the three Asters were knocked out simultaneously. We conclude that Asters participate in LDL cholesterol delivery from PM to ER, and their action depends in large part, but not exclusively, on PS. The data also indicate that PS participates in another delivery pathway, so far undefined, that is independent of Asters.

CIE XYZ Net: Unprocessing Images for Low-Level Computer Vision Tasks.

Cameras currently allow access to two image states: (i) a minimally processed linear raw-RGB image state (i.e., raw sensor data); or (ii) a highly-processed nonlinear image state (e.g., sRGB). There are many computer vision tasks that work best with a linear image state, such as image deblurring and image dehazing. Unfortunately, the vast majority of images are saved in the nonlinear image state. Because of this, a number of methods have been proposed to "unprocess" nonlinear images back to a raw-RGB state. However, existing unprocessing methods have a drawback because raw-RGB images are sensor-specific. As a result, it is necessary to know which camera produced the sRGB output and use a method or network tailored for that sensor to properly unprocess it. This paper addresses this limitation by exploiting another camera image state that is not available as an output, but it is available inside the camera pipeline. In particular, cameras apply a colorimetric conversion step to convert the raw-RGB image to a device-independent space based on the CIE XYZ color space before they apply the nonlinear photo-finishing. Leveraging this canonical image state, we propose a deep learning framework, CIE XYZ Net, that can unprocess a nonlinear image back to the canonical CIE XYZ image. This image can then be processed by any low-level computer vision operator and re-rendered back to the nonlinear image. We demonstrate the usefulness of the CIE XYZ Net on several low-level vision tasks and show significant gains that can be obtained by this processing framework. Code and dataset are publicly available at https://github.com/mahmoudnafifi/CIE_XYZ_NET.

A Little Bit More: Bitplane-Wise Bit-Depth Recovery.

Imaging sensors digitize incoming scene light at a dynamic range of 10-12 bits (i.e., 1024-4096 tonal values). The sensor image is then processed onboard the camera and finally quantized to only 8 bits (i.e., 256 tonal values) to conform to prevailing encoding standards. There are a number of important applications, such as high-bit-depth displays and photo editing, where it is beneficial to recover the lost bit depth. Deep neural networks are effective at this bit-depth reconstruction task. Given the quantized low-bit-depth image as input, existing deep learning methods employ a single-shot approach that attempts to either (1) directly estimate the high-bit-depth image, or (2) directly estimate the residual between the high- and low-bit-depth images. In contrast, we propose a training and inference strategy that recovers the residual image bitplane-by-bitplane. Our bitplane-wise learning framework has the advantage of allowing for multiple levels of supervision during training and is able to obtain state-of-the-art results using a simple network architecture. We test our proposed method extensively on several image datasets and demonstrate an improvement from 0.5dB to 2.3dB PSNR over prior methods depending on the quantization level.

Effects of feeding lubabegron on gas emissions, growth performance, and carcass characteristics of beef cattle housed in small-pen environmentally monitored enclosures during the last 3 mo of the finishing period.

The development of technologies that promote environmental stewardship while maintaining or improving the efficiency of food animal production is essential to the sustainability of producing a food supply to meet the demands of a growing population. As such, Elanco (Greenfield, IN) pursued an environmental indication for a selective ß-modulator (lubabegron; LUB). LUB was recently approved by the United States Food and Drug Administration (FDA) to be fed to feedlot cattle during the last 14 to 91 d of the feeding period for reductions in gas emissions/kg of unshrunk final BW and HCW. A 4 × 2 factorial arrangement of treatments was used with the factors of dose (0.0, 1.38, 5.5, or 22.0 mg·kg-1 DM basis) and sex (steers or heifers). Three 91-d cycles were conducted (112 cattle/cycle) with each dose × sex combination being represented by a single cattle pen enclosure (CPE; 14 cattle/CPE) resulting in a total of 168 steers and 168 heifers (n = 6 replicates/dose). There were no interactions observed between dose and sex for any variable measured in the study (P ≥ 0.063). Five gases were evaluated for all pens based on CPE concentrations relative to ambient air: NH3, CH4, N2O, H2S, and CO2. Cumulative NH3 gas emissions were reduced by feeding cattle 5.5 and 22.0 mg·kg-1 LUB (P ≤ 0.023) and tended (P = 0.076) to be lower for the cattle fed 1.38 mg·kg-1 LUB compared with the negative controls (CON). The cumulative NH3 gas emission reductions of 960 to 1032 g, coupled with HCW increases (P ≤ 0.019) of 15 to 16 kg for all LUB doses vs. CON, led to reductions in NH3 gas emissions/kg HCW for all three LUB treatments (P ≤ 0.004). Similar to HCW, reductions in NH3 gas emissions/kg of unshrunk final BW were observed for all LUB doses (P ≤ 0.009) and were attributable to both decreases in NH3 gas emissions and numerical increases in BW. Dose had no effect on cumulative emissions or emissions standardized by BW or HCW for the other four gases (P ≥ 0.268). LUB is a novel tool to reduce emissions of NH3 gas per kilogram of unshrunk live BW and hot carcass weight.

Mobile Computational Photography: A Tour.

The first mobile camera phone was sold only 20 years ago, when taking pictures with one's phone was an oddity, and sharing pictures online was unheard of. Today, the smartphone is more camera than phone. How did this happen? This transformation was enabled by advances in computational photography-the science and engineering of making great images from small-form-factor, mobile cameras. Modern algorithmic and computing advances, including machine learning, have changed the rules of photography, bringing to it new modes of capture, postprocessing, storage, and sharing. In this review, we give a brief history of mobile computational photography and describe some of the key technological components, including burst photography, noise reduction, and super-resolution. At each step, we can draw naive parallels to the human visual system.

Effects of voluntary removal of ractopamine hydrochloride (Optaflexx) on live performance and carcass characteristics of beef steers.

There is a lack of consistency across the globe in how countries establish tissue ractopamine residue limits and which residue limits are applied to various tissues, particularly for edible noncarcass tissues. Therefore, some U.S. beef slaughter organizations have recommended a 48-h voluntary removal of ractopamine before slaughter to meet residue requirements of specific export countries and maintain international trade. Our objective was to assess the impact of voluntary removal of ractopamine hydrochloride (Optaflexx; Elanco, Greenfield, IN) up to 8 d before slaughter on growth performance and carcass characteristics. Crossbred beef steers (60 pens of 10 animals/pen) with an initial shrunk body weight (BW) of 611.8 ± 10 kg SEM were fed one of six treatments over 42 d. Treatments included a control that did not receive ractopamine, on-label use of ractopamine (0-d withdrawal), and 2, 4, 6, or 8 d of voluntary removal of ractopamine from feed before slaughter. The start of ractopamine feeding (30.1 mg/kg of diet dry matter for 32 d) was staggered, so that blocks could be slaughtered on the same day. Dry matter intake was decreased by 0.5 kg/d when ractopamine was fed with a 0-d withdrawal (P = 0.04) compared with the control, but was not altered (P = 0.56) as the duration of ractopamine removal increased from 0 to 8 d. Final BW, total BW gain, and average daily BW gain were increased by feeding ractopamine with a 0-d withdrawal (P = 0.09) compared with the control, but these variables decreased in a linear manner (P = 0.10) as the duration of removal increased from 0 to 8 d. Gain efficiency was improved by 15% (P < 0.01) by feeding ractopamine with a 0-d withdrawal compared with the control, and gain efficiency decreased linearly (P = 0.06) as the duration of ractopamine removal increased. Approximately 2/3 of the increase in gain efficiency remained after 8 d of removal. Hot carcass weight was increased by 6 kg (P = 0.02) by feeding ractopamine with a 0-d withdrawal compared to the control. Measured carcass characteristics were not altered by ractopamine feeding or by removal before slaughter (P ≥ 0.24). The consequences of voluntary removal of ractopamine up to 8 d before slaughter were a linear decrease in live BW gain (0.64 kg/d), poorer gain efficiency, and numerically lighter carcass weight.

Scap structures highlight key role for rotation of intertwined luminal loops in cholesterol sensing.

The cholesterol-sensing protein Scap induces cholesterol synthesis by transporting membrane-bound transcription factors called sterol regulatory element-binding proteins (SREBPs) from the endoplasmic reticulum (ER) to the Golgi apparatus for proteolytic activation. Transport requires interaction between Scap's two ER luminal loops (L1 and L7), which flank an intramembrane sterol-sensing domain (SSD). Cholesterol inhibits Scap transport by binding to L1, which triggers Scap's binding to Insig, an ER retention protein. Here we used cryoelectron microscopy (cryo-EM) to elucidate two structures of full-length chicken Scap: (1) a wild-type free of Insigs and (2) mutant Scap bound to chicken Insig without cholesterol. Strikingly, L1 and L7 intertwine tightly to form a globular domain that acts as a luminal platform connecting the SSD to the rest of Scap. In the presence of Insig, this platform undergoes a large rotation accompanied by rearrangement of Scap's transmembrane helices. We postulate that this conformational change halts Scap transport of SREBPs and inhibits cholesterol synthesis.

Last step in the path of LDL cholesterol from lysosome to plasma membrane to ER is governed by phosphatidylserine.

Animal cells acquire cholesterol from receptor-mediated uptake of low-density lipoprotein (LDL), which releases cholesterol in lysosomes. The cholesterol moves to the endoplasmic reticulum (ER), where it inhibits production of LDL receptors, completing a feedback loop. Here we performed a CRISPR-Cas9 screen in human SV589 cells for genes required for LDL-derived cholesterol to reach the ER. We identified the gene encoding PTDSS1, an enzyme that synthesizes phosphatidylserine (PS), a phospholipid constituent of the inner layer of the plasma membrane (PM). In PTDSS1-deficient cells where PS is low, LDL cholesterol leaves lysosomes but fails to reach the ER, instead accumulating in the PM. The addition of PS restores cholesterol transport to the ER. We conclude that LDL cholesterol normally moves from lysosomes to the PM. When the PM cholesterol exceeds a threshold, excess cholesterol moves to the ER in a process requiring PS. In the ER, excess cholesterol acts to reduce cholesterol uptake, preventing toxic cholesterol accumulation. These studies reveal that one lipid-PS-controls the movement of another lipid-cholesterol-between cell membranes. We relate these findings to recent evidence indicating that PM-to-ER cholesterol transport is mediated by GRAMD1/Aster proteins that bind PS and cholesterol.

Learning Raw Image Reconstruction-Aware Deep Image Compressors.

Deep learning-based image compressors are actively being explored in an effort to supersede conventional image compression algorithms, such as JPEG. Conventional and deep learning-based compression algorithms focus on minimizing image fidelity errors in the nonlinear standard RGB (sRGB) color space. However, for many computer vision tasks, the sensor's linear raw-RGB image is desirable. Recent work has shown that the original raw-RGB image can be reconstructed using only small amounts of metadata embedded inside the JPEG image [1]. However, [1] relied on the conventional JPEG encoding that is unaware of the raw-RGB reconstruction task. In this paper, we examine the ability of deep image compressors to be "aware" of the additional objective of raw reconstruction. Towards this goal, we describe a general framework that enables deep networks targeting image compression to jointly consider both image fidelity errors and raw reconstruction errors. We describe this approach in two scenarios: (1) the network is trained from scratch using our proposed joint loss, and (2) a network originally trained only for sRGB fidelity loss is later fine-tuned to incorporate our raw reconstruction loss. When compared to sRGB fidelity-only compression, our combined loss leads to appreciable improvements in PSNR of the raw reconstruction with only minor impact on sRGB fidelity as measured by MS-SSIM.

Improved tagged cardiac MRI myocardium strain analysis by leveraging cine segmentation.

BACKGROUND AND OBJECTIVES: Tagged MR images provide an effective way for regional analysis of the myocardium strain. A reliable myocardium strain analysis requires both correct segmentation and accurate motion tracking of the myocardium during the cardiac cycle. While many algorithms have been proposed for accurate tracking of the myocardium in tagged MR images, little focus has been placed on ensuring correct segmentation of the tagged myocardium during the cardiac cycle. Myocardial strain analysis is usually done by segmenting the myocardium in end-diastole, generating a mesh from the segmentation, propagating the mesh through the cardiac cycle using the output deformation field from motion tracking, and measuring strain on the deforming mesh. Due to the imposed tag strips on the anatomy, identification of the myocardium boundaries is challenging in tagged MR images. As a result, there is no guarantee that the propagated mesh is annotating the myocardium accurately through the cardiac cycle. Moreover, clinical studies indicate that incorrect myocardium annotation can result in overestimation of myocardial strains. METHODS: We introduce a method to improve reliability of strain analysis by proposing a mesh which correctly segments the myocardium in tagged MRI by leveraging the available cine MRI segmentation. In particular, we generate a series of mesh proposals using the cine MRI segmentation and find the propagated mesh proposal which gives the most accurate full-cycle myocardium segmentation. RESULTS: The mesh selection algorithm was tested on 22 2D MRI scans of diseased and healthy hearts. The proposed algorithm provided more accurate whole-cycle myocardium segmentation compared to the propagated end-diastolic mesh. Regional myocardium strain was measured for 10 3D MRI scans of healthy volunteers using the proposed mesh and the end-diastolic mesh. The measured strain using the proposed mesh was more similar to the expected myocardium strain for a healthy heart than the measured strain using the end-diastolic mesh. CONCLUSION: The proposed approach provides accurate whole-cycle tagged myocardium segmentation and more reliable myocardium strain analysis.

Closed rupture of the flexor tendon pulleys of the thumb.

Closed rupture of the thumb flexor tendon pulleys is rare and, to our knowledge, the MRI findings associated with this injury have not been described in the radiology literature. The injury can result in nonspecific clinical and physical exam findings and advanced imaging may be crucial to diagnosis. A familiarity with the normal anatomy of the flexor mechanism of the thumb and the imaging appearance of pathology is therefore necessary. We report a case of a 31-year-old female nurse with ruptures of several of the thumb flexor pulleys and the corresponding findings on MRI.

Growth hormone acts on liver to stimulate autophagy, support glucose production, and preserve blood glucose in chronically starved mice.

When mice are subjected to 60% calorie restriction for several days, they lose nearly all of their body fat. Although the animals lack energy stores, their livers produce enough glucose to maintain blood glucose at viable levels even after a 23-hour fast. This adaptation is mediated by a marked increase in plasma growth hormone (GH), which is elicited by an increase in plasma ghrelin, a GH secretagogue. In the absence of ghrelin, calorie-restricted mice develop hypoglycemia, owing to diminished glucose production. To determine the site of GH action, in the current study we used CRISPR/Cas9 and Cre recombinase technology to produce mice that lack GH receptors selectively in liver (L-Ghr-/- mice) or in adipose tissue (Fat-Ghr-/- mice). When subjected to calorie restriction and then fasted for 23 hours, the L-Ghr-/- mice, but not the Fat-Ghr-/- mice, developed hypoglycemia. The fall in blood glucose in L-Ghr-/- mice was correlated with a profound drop in hepatic triglycerides. Hypoglycemia was prevented by injection of lactate or octanoate, two sources of energy to support gluconeogenesis. Electron microscopy revealed extensive autophagy in livers of calorie-restricted control mice but not in L-Ghr-/- mice. We conclude that GH acts through its receptor in the liver to activate autophagy, preserve triglycerides, enhance gluconeogenesis, and prevent hypoglycemia in calorie-restricted mice, a model of famine.

As-projective-as-possible bias correction for illumination estimation algorithms.

Illumination estimation is the key routine in a camera's onboard auto-white-balance (AWB) function. Illumination estimation algorithms estimate the color of the scene's illumination from an image in the form of an R, G, B vector in the sensor's raw-RGB color space. While learning-based methods have demonstrated impressive performance for illumination estimation, cameras still rely on simple statistical-based algorithms that are less accurate but capable of executing quickly on the camera's hardware. An effective strategy to improve the accuracy of these fast statistical-based algorithms is to apply a post-estimate bias-correction function to transform the estimated R, G, B vector such that it lies closer to the correct solution. Recent work by Finlayson [Interface Focus8, 20180008 (2018)2042-889810.1098/rsfs.2018.0008] showed that a bias-correction function can be formulated as a projective transform because the magnitude of the R, G, B illumination vector does not matter to the AWB procedure. This paper builds on this finding and shows that further improvements can be obtained by using an as-projective-as-possible (APAP) projective transform that locally adapts the projective transform to the input R, G, B vector. We demonstrate the effectiveness of the proposed APAP bias correction on several well-known statistical illumination estimation methods. We also describe a fast lookup method that allows the APAP transform to be performed with only a few lookup operations.

Lysosomal cholesterol export reconstituted from fragments of Niemann-Pick C1.

Niemann-Pick C1 (NPC1) is a polytopic membrane protein with 13 transmembrane helices that exports LDL-derived cholesterol from lysosomes by carrying it through the 80 Å glycocalyx and the 40 Å lipid bilayer. Transport begins when cholesterol binds to the N-terminal domain (NTD) of NPC1, which projects to the surface of the glycocalyx. Here, we reconstitute cholesterol transport by expressing the NTD as a fragment separate from the remaining portion of NPC1. When co-expressed, the two NPC1 fragments reconstitute cholesterol transport, indicating that the NTD has the flexibility to interact with the remaining parts of NPC1 even when not covalently linked. We also show that cholesterol can be transferred from the NTD of one full-length NPC1 to another NPC1 molecule that lacks the NTD. These data support the hypothesis that cholesterol is transported through interactions between two or more NPC1 molecules.

BHLHE40, a third transcription factor required for insulin induction of SREBP-1c mRNA in rodent liver.

In obesity, elevated insulin causes fatty liver by activating the gene encoding SREBP-1c, a transcription factor that enhances fatty acid synthesis. Two transcription factors, LXRα and C/EBPß, are necessary but not sufficient for insulin induction of hepatic SREBP-1c mRNA. Here, we show that a third transcription factor, BHLHE40, is required. Immunoprecipitation revealed that BHLHE40 binds to C/EBPß and LXRα in livers of rats that had fasted and then refed. Hepatic BHLHE40 mRNA rises rapidly when fasted rats are refed and when rat hepatocytes are incubated with insulin. Preventing this rise by gene knockout in mice or siRNAs in hepatocytes reduces the insulin-induced rise in SREBP-1c mRNA. Although BHLHE40 is necessary for insulin induction of SREBP-1c, it is not sufficient as demonstrated by failure of lentiviral BHLHE40 overexpression to increase hepatocyte SREBP-1c mRNA in the absence of insulin. Thus, an additional event is required for insulin to increase SREBP-1c mRNA.